Author Topic: Input sensitivity question  (Read 922 times)

0 Members and 1 Guest are viewing this topic.

Offline n2qew

  • Member
  • *
  • Posts: 49
Input sensitivity question
« on: August 24, 2014, 11:45:22 PM »
New user here, ordered and waiting for hardware.
But, that didn't stop me from building an antenna ahead of time. dual square loops, 3/4" PVC, 40 inches on a side, 7 turns of 18 gauge per. Totally weather-tight - hey, it's PVC conduit.

Anyway, I just had to do some testing, and given I don't have the hardware yet, I had to make due. Ended up running the two loops into an ADC running 24 bits@ 48khz. Not the same resolution as the real hardware, but the traces look similar. Loops were terminated into 600 ohms vs. the 2.4k on the real amp.

What I'm wondering is, what sort of signal level is expected at the input of the real amp for strokes not really close? I see some that I estimate as being about 400 miles away based on what was going on on the map at the time, as well as a few that were closer, like 25 miles. On the attached screen shots, the -1db point is about 8mv from zero. Tough to do db to volt conversions with pulses, but....

I was pleased that it appears that there is no high level magnetic noise here - 23 kv single phase running by the front of the house, and the traces attached are at ground level about 15 feet away from the line - not where I will be putting it the loops permanently anyway. what little 60hz stuff there was was easily removed with a 2khz high pass filter, which will apparently be done by the real amp.

I'm really looking forward to getting online. I am near Wilmington, NC - dead center between the stations in MD and FL on the East Coast.


Offline JonathanW

  • Engineer
  • Forecaster
  • *****
  • Posts: 323
Re: Input sensitivity question
« Reply #1 on: August 25, 2014, 08:07:54 AM »
I'm very much looking forward to increased coverage down the East Coast - welcome!

And your question regarding signal levels is a good one.  I've tried to find some information on this without a lot of success.

However, using my BO system as a reference, for the H-field I use the (now pretty standard) close-wound, 300 mm x 7.5 mm ferrites with roughly 700 turns of AWG 26 wire on 2000-mu cores.  For far (middle of the country) strikes, I'll typically run the system with a total voltage gain of 2560, and strikes clock in at between 50 mVpp and 3 Vpp.

When strikes get closer, I'll dial the total gain down to as little as 200 to keep the most powerful signals below clipping (again, about 3 Vpp).  Even then, with a really close strike the signal may clip.

I don't know whether that helps - I'm not sure what signal levels are typically pulled off of air-core loops.  However, I'm pretty sure there are other stations on that have that type of H-field antenna (e.g. Panama City, FL; Rogers City, MI; Foley, MN; Appleton, WI -- you can find more details for the stations here: http://www.lightningmaps.org/blitzortung/america/index.php?bo_page=statistics&bo_show=station&lang=en), and you can see their signal levels and gains posted at the link in the "Signal Stations - Tool - Region 3" thread.

Best,
Jonathan
Station 1013, Montgomery Village, MD

Offline miraculon

  • Sunrise Side Weather
  • Forecaster
  • *****
  • Posts: 4107
  • KE8DAF
    • Sunrise Side Weather in Rogers City MI USA
Re: Input sensitivity question
« Reply #2 on: August 25, 2014, 08:39:20 AM »
Quote
I don't know whether that helps - I'm not sure what signal levels are typically pulled off of air-core loops.  However, I'm pretty sure there are other stations on that have that type of H-field antenna (e.g. Panama City, FL; Rogers City, MI; Foley, MN; Appleton, WI

My station #706 is the one with the 1 meter loops (9 turns of 20AWG).
The other station #668 has ferrite rods.

Greg H.



Blitzortung Stations #706 and #1682
CoCoRaHS: MI-PI-1
CWOP: CW4114 and KE8DAF-13
WU: KMIROGER7
Amateur Radio Callsign: KE8DAF

Offline JonathanW

  • Engineer
  • Forecaster
  • *****
  • Posts: 323
Re: Input sensitivity question
« Reply #3 on: August 25, 2014, 08:56:26 AM »
Quote
I don't know whether that helps - I'm not sure what signal levels are typically pulled off of air-core loops.  However, I'm pretty sure there are other stations on that have that type of H-field antenna (e.g. Panama City, FL; Rogers City, MI; Foley, MN; Appleton, WI

My station #706 is the one with the 1 meter loops (9 turns of 20AWG).
The other station #668 has ferrite rods.

Greg H.

How do you think the two compare?

Offline n2qew

  • Member
  • *
  • Posts: 49
Re: Input sensitivity question
« Reply #4 on: August 25, 2014, 11:01:44 AM »
"and you can see their signal levels and gains posted at the link in the "Signal Stations - Tool - Region 3" thread"

Is the value displayed what the ADC interprets it to be after all gain and amps? How does the gain work? If the gain is set to say 2000, and the value displayed is 2 volts, does that mean the input level to the first stage is 1 mv P-P? Or, is there other math involved?

Also, does the system look at how often clipping happens, and push the gain up or down, or is it something that we'll need to adjust. If manual, would having two H-field loops and amps help - a low gain one for closer activity, and a higher gain one for distant activity?

Offline miraculon

  • Sunrise Side Weather
  • Forecaster
  • *****
  • Posts: 4107
  • KE8DAF
    • Sunrise Side Weather in Rogers City MI USA
Re: Input sensitivity question
« Reply #5 on: August 25, 2014, 11:48:45 AM »
Quote
I don't know whether that helps - I'm not sure what signal levels are typically pulled off of air-core loops.  However, I'm pretty sure there are other stations on that have that type of H-field antenna (e.g. Panama City, FL; Rogers City, MI; Foley, MN; Appleton, WI

My station #706 is the one with the 1 meter loops (9 turns of 20AWG).
The other station #668 has ferrite rods.

Greg H.

How do you think the two compare?

It is difficult to tell, because one is on "Red" with full automatic control. It seems to gain down the H-Field loops and rely more on E-Field.

The "green" ( USB6.8 ) is highly modified with HPF set for the higher range, 10K terminator resistors, single-ended input and an adjustable threshold. I am running the highest gain setting on the gain jumpers.

If they were both on the same type of system, I could give you a better answer. I have detected strikes on the ferrites/green that Red missed and vice-versa. I get good range out of both systems. Having said that, the participation rates are lower on the old "green/ferrites" system. If anything, I attribute this to E-Field. Before I built and installed the E-Field, the results were closer between the two systems.

Greg H.



Blitzortung Stations #706 and #1682
CoCoRaHS: MI-PI-1
CWOP: CW4114 and KE8DAF-13
WU: KMIROGER7
Amateur Radio Callsign: KE8DAF

Offline JonathanW

  • Engineer
  • Forecaster
  • *****
  • Posts: 323
Re: Input sensitivity question
« Reply #6 on: August 25, 2014, 01:06:03 PM »
"and you can see their signal levels and gains posted at the link in the "Signal Stations - Tool - Region 3" thread"

Is the value displayed what the ADC interprets it to be after all gain and amps? How does the gain work? If the gain is set to say 2000, and the value displayed is 2 volts, does that mean the input level to the first stage is 1 mv P-P? Or, is there other math involved?

Also, does the system look at how often clipping happens, and push the gain up or down, or is it something that we'll need to adjust. If manual, would having two H-field loops and amps help - a low gain one for closer activity, and a higher gain one for distant activity?

To answer the second question first...unless you set the "auto adapt to noise" (or go full auto), the system doesn't do any adjustment of gain or threshold - it just goes with what you set (auto amplitude filter tries to intelligently filter out repetitive signals with the same amplitude).

Actually, much of the talk with multiple amps for the system has revolved around both an H-field and an E-field filter - H-field for farther away, and E-field for closer up (though both can pick up far signals with high gain set).  Right now, if one amp's signals exceed interference mode thresholds, the whole system goes into interference mode.  My understanding is that this will eventually change, and there will be two interference modes (one for each amp), at which point the "farther out/closer in" scenario will be viable.

As for whether what is displayed reflects actual signal levels - the answer is, I don't know.  I looked at that a few months back, and I thought the discovery board's microcontroller might have an A/D system that topped out at 3.3 V.  This compares well with comments from the developers that +/- 1.5 V or so will drive the system into clipping.